Phil Zarrow: Brook, we talk about choosing a flux. We start off with these classifications. What is the meaning of it?
Brook Sandy-Smith: They can mean a lot of different things to a lot of different people in my book, but you start out with the flux basis.
Phil Zarrow: Right.
Brook Sandy-Smith: That's RO, RE, OR—the first two letters of the classification.
Then the third letter of the classification instructs you as to the activity of the flux or the reliability of the flux. That's characterized with several different and test methods, which we'll talk about just in a short minute.
Then you have a zero or one at the end, which determines whether there are zero halides or there are halides present. This halide content actually, also, if it's a zero is just a zero, and it will contribute an L to the L, M, or H rating. If there are higher levels of Ionics in the raw flux, that can also contribute an M or an H to the classification.
Swinging back around to the test methods that influence the L, M, or H rating, there are five different test methods. They all assess the interaction between the four different factors for electrochemical migration, which are humidity, metal, ionic contamination, and voltage.
If you look at the first two, copper mirror and copper corrosion, those both look at the interaction between the ionics from the flux and the copper that's present. For copper mirror, it's a very thin layer of copper and you look at how the raw flux interacts with the copper and whether it removes it from the surface.
With copper corrosion, you look at flux residue, so the flux has been heated and then you look at how the copper coupon changes when it's put into a human environment. That's another way to look at the interaction between the flux and the copper.
Halide content is measured by ion chromatography. Not only are you looking at the level of ions in the raw flux, but you're also looking at the species. Ion chromatography will give you a spectrum with different peaks, which you can determine which peaks are from which elements. You can see the source of the ions.
We have SIR and ECM. SIR is surface insulation resistance test, whereas the ECM is electrochemical migration test. They're both very similar in that they use the same kind of coupon design, with an interlaced finger pattern.
Phil Zarrow: The comb, the infamous comb.
Brook Sandy-Smith: The comb.
Phil Zarrow: Yeah.
Brook Sandy-Smith: Where you have a positive side and a negative side interlaced and then you create a voltage in between them. These coupons are then placed into the environmental chamber and exposed to temperature and humidity, which will accelerate the potential for dendritic growth, and that's a standardized way that you can test the way that the flux residue interacts with the circuitry.
Phil Zarrow: Okay. Here we're talking about tests that are standardized and very idealized too. What about the real test of the assembly in practice by the practitioner? Say you're qualifying the process and controlling the process. Where were you looking at here?
Brook Sandy-Smith: Well, process control is a whole different animal. You can't do SIR on every assembly because there isn't a little coupon on the corner that you can break out and seven days that you can wait on every assembly to decide whether it's reliable or not. The industry has really yearned for a test method that's quick and easy and can be done on every assembly.
A lot of times the process control method that we've used is ROSE, the ROSE test method, which is resistivity of solvent extract. Typically, this is done in a bag extraction so you'll put a whole plastic bag around the whole assembly; add solvent; shake it around, or however, to bring the solvent in contact with all of the residues that are on the entire assembly. Then take the resulting solvent, and pass current through it to see how conductive the solvent is. That's a measure of the ionic cleanliness of your port.
Phil Zarrow: Right. Of course, we have machines for doing this too, besides the old plastic bag but-
Brook Sandy-Smith: Of course.
Phil Zarrow: Right.
Brook Sandy-Smith: You can see how it's kind of time consuming and you're using a lot of solvent for a small amount of contamination. There's been the question over the years, how accurate could this really be? What if you have contamination on one part of the board?
Phil Zarrow: Exactly. It's giving you a global look at the contamination but with regard to failure analysis, process failure analysis, how do we pin point? What would be a localized methodology to use?
Brook Sandy-Smith: Well, one method that's been developed recently, which is often referred to as localized extraction followed by IC, or ion chromatography, that's the C3 method, which has a nozzle, which comes down in contact with the board and it's pliable so it can conform over the top of components, or a lot of times will focus on the corner of a component. Then it uses cool steam to extract any ionic contaminants on that small area. Then, the nozzle sucks up the extract and checks the resistivity, just like the ROSE method.
Phil Zarrow: Right.
Brook Sandy-Smith: Also, that extract can be collected and put through IC so that you can use the ion chromatography to identify the species or the elements that were there in the ionic contamination. That makes it easier to trace it back to a part of your process.
Phil Zarrow: I guess the thing to understand is, we want our listeners to understand is, there's not one test that is all conclusive. Essentially, you're looking at several or a combination of tests to really give you the full picture, at least to date right now.
Brook Sandy-Smith: Of course, and the requirements are going to be different based on the assembly that you're talking about, based on the longevity that it's expected to have in the field, based on the kind of environment that it's supposed to be in, whether it's conformally coated, all sorts of things like that. These different tools that are emerging technologies, are really useful in assessing small areas of concern. It's a quick test that you can use for process control, and test the same spot on every assembly to see changes over time.
When you have your spec-ed in, end of the design process, this is the one we're going with assembly, you can make sure that you don't exceed that level of contamination in the future. When you see changes, that's when you go back to your process and try to figure out what change—Is the reflow profile no longer as hot? Did I turn on nitrogen and change the way that the air is flowing the machine?
Phil Zarrow: Right.
Brook Sandy-Smith: There are all sorts of different things that can change.
Phil Zarrow: Right.
Brook Sandy-Smith: These are really great tools to help you go back to your process and figure out what's changed.
Phil Zarrow: Right. How does J-STD-001 look at these particular advances?
Brook Sandy-Smith: I mean, in the J-STD-001, looking at the cleanliness of your circuit board, they propose the ROSE method, but also leave it open for the user and supplier to agree upon different test methods that might also be acceptable to assess this. They really want a process control where you're making sure the same level of cleanliness is consistent every time. Not necessarily that you're reaching a limit and now it's bad.
Phil Zarrow: Right. Right.
Brook Sandy-Smith: That's the difference between process control and flux classification. Flux classification has limits. It's standardized. It's telling you, this meets these expectations that you were having of a flux.
Phil Zarrow: Right.
Brook Sandy-Smith: Whereas process control is making sure that you're still processing everything correctly, and making sure that there aren't unforeseen changes from your materials when they're incoming or from your process or from your handling, that could affect the electrochemical reliability.
Phil Zarrow: A steady, consistent, controlled process.
Brook Sandy-Smith: Yes.
Phil Zarrow: Exactly. Brook, you've authored papers on this. Where can we find them?
Brook Sandy-Smith: You can find them right at www.indium.com or if you have questions, you can contact me directly as bsandy@indium.com.
Phil Zarrow: Thank you for your work, appreciate it.